S-EMBER - Multiple Choice: leaderboard
Metric: Accuracy (%; five options with LLM-written distractors drawn from models' wrong answers) (S-EMBER: 9,448 questions on 3,141 egocentric Ray-Ban Meta videos, answered from the stream up to the question's trigger time; open models at 128 uniformly sampled frames, proprietary models at their platform limits). Source: arxiv.org. Saturation forecast: Around 2028. 5 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 54.41 |
| 2 | Qwen 3 VL 32B Instruct | 42.4 |
| 3 | GPT-5.4 | 41.82 |
Interactive version: theaggregate.ai/benchmark?slug=s-ember-multiple-choice · How It Works · Data refreshed daily, snapshot 2026-09-26.